Limiting max filter capacitance on 1st stage of Bass amp B+, or Pi filter?

[satsloader]

Hi all,

I'm building a 2xKT88 UL bass amp. The PT is a MM P4550JT-G3, 510vdc B+ ( PT was given to me for servicing a couple of amps for a friend ). The OT is a A431S Dynaco clone ( Freaking huge! )

I emailed MM asking for a recommendation regarding a safe max amount of filtering on the first stage ( UL, so both the plates and screens ), but hadn't received a reply, so I thought I'd ask you fine folks!

I originally planned 200uf or 300uf for the first stage, 100uf for the phase inverter, 50uf each for the remaining two stages. ( I'm using ARS Marshall style can caps, and the chassis is Weber's Marshall 100 watt clone ).

Then I saw some warnings regarding putting too much capacitance on the first node, and a suggestion for using a Pi filter in place of the large 1st node.

Obviously, if going the PI filter route, I'd want to keep the choke's resistance as low as possible.. I have a Hammond 193H, rated for 5H@200ma., with 65 ohms of resistance. Does anyone feel 65ohms is too high? The configuration would be; Full wave recto, 100uf/5H/200uf, then plates/screens.. then as described above for the rest of the supply..

The only real downside I can see to going the PI route is finding a place for the choke ( The Dyna OT really is big, taking up most of the mounting area of normal OT and choke ).

I should mention this amp will be used in jazz/pop small venue situations. High volume is not a priority.

Any thoughts or suggestions?

Thanx!

Bill

[loudthud]

The choke in the original Dynaco amps was something like 1.5Hy 50 ohms. The 193H is not the best choice. A 159V or 159Y would work better with the caps you purpose. I've found that 100uF on the input and 200uF on the output side works well on bass amps. With less than 1Hy, the 120Hz will be less than 2V p-p on the output. The 200uF gives a nice low impedance down to 40Hz.

[Enzo]

You could also simplify the pi filter to a resistor instead of a choke.

What kind of rectifier you using? This is not my area of expertise, but I believe the cap size limit is for the rectifier tube rather than the transformer.

[loudthud]

Duncan's PSU designer is a great tool for evaluating what changes to a power supply do. As you increase the first capacitor, beyond a certain point you'll see that the voltage goes down and the RMS transformer current goes way up as does ripple current in the cap. You have to go round and round a couple of times to strike a balance between expectations and the parts you have available or are will to pay for.

It's important to get good numbers on the transformer for the program to work accurately. You need to know the open circuit secondary voltage and it's resistance. This is not the same as the advertised voltage which is usually quoted under load.

Once you get familiar with the program, you'll be amazed at the wonders a choke can do in a power supply.

[satsloader]

Thanx LT and Enzo,

I downloaded the PSU designer a long time ago, the first version if memory serves correctly, and couldn't make heads or tails of it, since I didn't know what to input properly in regards to values. I'll have a go at it again with a new download and see if there's a help file, or other support info on the net for it.

Enzo, if my understanding is correct, and I think LT's his last reply eludes to this.. As the first node's capacitance is increased beyond a certain point the stability of the PT becomes less, as though it can't fully supply the peaks.

I was hoping to get the official dirt from MM, but since I'm using their PT in a situation it wasn't specifically designed for, I'm guessing they don't want to touch it.. I don't blame them.

One of my concerns that I should have made clear in my first post is limiting the series resistance while succesfully de-coupling the pi filters enough that the PT doesn't suffer. The PSUD is probably my best bet in figuring that out, so I'll give it a go..

Thanx again,

Bill

[satsloader]

Sorry Enzo, I forgot to include, the recto is SS full wave.

Bill

[bob p]

I was hoping to get the official dirt from MM, but since I'm using their PT in a situation it wasn't specifically designed for, I'm guessing they don't want to touch it.. I don't blame them.

I don't think that your experience is unique. IME MM doesn't like to give out any specs on their iron, under any circumstances.

[bob p]

I downloaded the PSU designer a long time ago, the first version if memory serves correctly, and couldn't make heads or tails of it, since I didn't know what to input properly in regards to values. I'll have a go at it again with a new download and see if there's a help file, or other support info on the net for it.

PSUD now has a library of downloadable sample files, including models of complete amp circuits. That should help to get you started.

[Steve Conner]

Yes, you are right, the 1st filter capacitance limit is mainly because of tube rectifiers. Eventually you might reach a value that will overheat the PT due to the RMS-to-average current ratio getting too high, but I don't think 200 or even 300uF will do that.

Having said that, going to a Pi filter is a much better way of reducing ripple than just making the first capacitor bigger.

[Earl]

I asked MM for the specs on their 'brown Princeton' PT, and they actually sent me a data sheet...for whatever that's worth.

I'm not a PS designer, but I remember a couple of rules of thumb. One being that the discharge rate ought to be roughly 3 to 5 times the charge rate for decent ripple percent.
With chokes, you should make sure there's enough bias current flowing through them to satisfy their 'keep alive' requirement.
Make sure that the inrush current doesn't last long enough to toast the diodes/PT secondary.

[satsloader]

Hi Earl,

Regarding a choke's minimum current required in the circuit, isn't that if the choke is the first component after the rectifier, in other words, a true "choke input" power supply, instead of a Pi filter?

Work and HoneyDo's have kept me from exploring PSUD2.. I did look at Hammond's offerings again ( bummed I can't use the 193H, got two of them collecting dust and hate to buy another choke ). If getting another I'm looking at the 193K. 2.6H, 300ma., and only 21 ohms..

Bill

[bob p]

Yes, you are right, the 1st filter capacitance limit is mainly because of tube rectifiers. Eventually you might reach a value that will overheat the PT due to the RMS-to-average current ratio getting too high, but I don't think 200 or even 300uF will do that.

Having said that, going to a Pi filter is a much better way of reducing ripple than just making the first capacitor bigger.

there are lots of SS designs that use HUGE filter caps in the first position. one way to get around the RMS/Average current thing is to have a switchable limiting resistor/ soft start feature. then the problem goes away, right?

i have done a bit of modelling with asymmertric Pi filters where the first cap is smaller and the second cap is bigger. its always turned out to be a bad recipe, for a couple of reasons. first, the ability of the PSU filter to respond to a sudden huge current tap is impaired by the small size of the first cap. whether this is musically relevant... well, that's another question. second, the math shows that the best pi filters tend to be ones that are symmetrical. IIRC there's specific mention of this in the radiotron designer's handbook section on pi filters.

my preference, then, is SS rectification with a big Pi filter, and using a switch-out resistor for soft starting. am i missing something?

[Jag]

I've been using MM iron for a couple of years. I've always found them very helpful but I always call as opposed to email. Maybe try calling and ask for Paul. (818) 998-7791.

[Earl]

Sorry satsloader - I thought this was more or less the Fender style cap/choke deal.